Recently, unresolved hard (20-40 keV) X-ray emission has been discovered within the central 10 pc of the Galaxy, possibly indicating a large population of intermediate polars (IPs). Chandra and XMM-Newton measurements in the surrounding ~50 pc imply a much lighter population of IPs with $langle M_{rm WD} rangle approx 0.5 M_odot$. Here we use broad-band NuSTAR observations of two IPs: TV Columbae, which has a fairly typical but widely varying reported mass of $M_{rm WD} approx 0.5-1.0 M_odot$, and IGR J17303-0601, with a heavy reported mass of $M_{rm WD} approx 1.0-1.2 M_odot$. We investigate how varying spectral models and observed energy ranges influence estimated white dwarf mass. Observations of the inner 10 pc can be accounted for by IPs with $langle M_{rm WD} rangle approx 0.9 M_odot$, consistent with that of the CV population in general, and the X-ray observed field IPs in particular. The lower mass derived by Chandra and XMM-Newton appears to be an artifact of narrow energy band fitting. To explain the (unresolved) CHXE by IPs requires an X-ray (2-8 keV) luminosity function (XLF) extending down to at least $5times10^{31}$ erg/s. The CHXE XLF, if extended to the surrounding ~50 pc observed by Chandra and XMM-Newton, requires at least ~20-40% of the $sim$9000 point sources are IPs. If the XLF extends just a factor of a few lower in luminosity, then the vast majority of these sources are IPs. This is in contrast to recent observations of the Galactic ridge, where the bulk of the 2-8 keV emission is ascribed to dwarf novae.